Mind Map HTML Output

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The workflow ingests the user’s free-text/markdown in step 1 (“Collect / parse the content”) and embeds it into the first-party standalone HTML template by replacing __MINDMAP_DATA__ in step 3, which the browser then renders via Mind Elixir.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The generated HTML imports and executes remote assets from the jsDelivr CDN at runtime (e.g., https://cdn.jsdelivr.net/npm/mind-elixir@5/dist/MindElixir.js, https://cdn.jsdelivr.net/npm/mind-elixir@5/dist/MindElixir.css, and https://cdn.jsdelivr.net/npm/mind-elixir@5/dist/PlaintextConverter.js), so these external URLs are required runtime dependencies that run remote code in the client.